Dynamics of recharge, evapotranspiration, and water surplus of a large-scale lowland catchment with respect to changes in climate and land use 

Abstract

The Lower Spree Catchment located in Berlin-Brandenburg, Germany, poses a vital tributary of the Elbe river and lifeline of the UNESCO Biosphere Reserve Spree Forest. The region is of high agricultural and economic importance which, exasperated by the consequences of climate change, facilitates dropping groundwater levels by 2-3 m over the last 20 years, resulting in a region with increasing water stress. This research assesses recharge dynamics of the Lower Spree catchment via the utilization of the Soil and Water Assessment Tool (SWAT) to enable an integrated water resource management for the area. Key to prospectively managing our groundwater resources in a more sustainable way is to understand recharge dynamics and to use this information for water management strategies. This will help to assess stored and refreshed water volumes in the subsurface and to understand how they will respond to shifts in climate, human activities, and land use change. Furthermore, accurately quantifying surplus water volumes, such as run-off during wet periods, will enable us to calculate water volumes for a better adaption to climate change, e.g. implementation of artificial groundwater recharge. For this purpose, the presented study integrated relevant pre-processed hydroclimatic and geospatial datasets into a hydrological model that was subsequently simulated for a period of 50 years (1971-2020). Calibration was conducted with the SWAT Calibration and Uncertainty Programs (SWAT-CUP) against the recorded daily streamflow and monthly evapotranspiration (MOD16A2 - MODIS/Terra Net Evapotranspiration 8-Day L4 Global 500m SIN Grid) data for the periods of 1980-2020 and 2000-2020 respectively, revealing a significant correlation between multisite daily observed streamflow and actual evapotranspiration at a monthly resolution. Model performance has finally been evaluated by calculating the water balance of the Lower Spree Catchment and estimating groundwater recharge. The model calculates a mean annual groundwater recharge of 77 mm for the simulated period and an overall change in soil water storage of -83 mm annually. Spatial analysis highlights varying recharge rates, with urban regions exhibiting lower rates of 0-9 mm of daily groundwater recharge compared to 19-32 mm in forested and irrigated areas, likely due to a higher degree of surface sealing in urban regions. Statistical analysis using modified M-K tests demonstrate seasonal fluctuations in groundwater recharge, where a Kendall’s tau coefficient of 0.07 supported by a p-value of 0.002 indicates an increase in groundwater recharge in winter. In spring and summer tau coefficients of -0.12 and -0.15 and p-values of 0.0049 and 9.45E-05 respectively show a decrease in groundwater recharge; no definitive patterns were found for autumn. We are planning to improve the recharge calculations through the incorporation of tile drainage and detailed irrigation data. Additionally, high resolution soil moisture and discharge field-measurements from the model region will provide further validation metrics via comparison of the field data with the simulated surface run-off, interflow, and baseflow.